Jakarta, INTI – In recent years, China has demonstrated a significant shift in its approach to artificial intelligence (AI) by increasingly adopting open-source models. This trend is seen as a strategic move to accelerate innovation and broaden AI adoption, with many calling it the "Android moment" of artificial intelligence.
DeepSeek and the Changing AI Landscape in China
Leading this shift is Chinese AI startup DeepSeek, which released its R1 model earlier this year. This model challenges the dominance of U.S.-based AI technology and raises questions about the effectiveness of large-scale investments by global tech companies in developing large language models (LLMs) and data centers.
DeepSeek's success highlights how an open-source strategy can drive innovation and enable wider adoption. Wei Sun, a lead AI analyst at Counterpoint Research, noted that many companies have already implemented this model, forcing tech giants like Baidu to start open-sourcing their AI models.
Baidu’s Strategic Shift and Major Industry Players
On March 16,2025, Baidu released the latest version of its AI model, Ernie 4.5, along with a new reasoning model, Ernie X1, free for individual users. Additionally, Baidu also plans to open-source the Ernie 4.5 model series by the end of June. This marks a shift from a proprietary licensing business model to a more inclusive open-source approach.
This move surprised many, as Baidu previously strongly supported a proprietary business model and resisted open-source initiatives. However, with DeepSeek's success, more Chinese companies are adopting similar strategies to stay competitive.
Open-Source vs. Proprietary Models
Generally, open-source refers to software whose source code is freely available for modification and redistribution. While AI models such as Meta’s Llama and Google’s Gemma has been released, many experts argue that these models are not entirely open due to restrictive licenses and undisclosed training data.
In contrast, DeepSeek R1 is distributed under the MIT license, one of the most permissive and widely used open-source licenses, allowing for commercial use. DeepSeek has even launched an "Open-Source Week" to share more technical details about its model development.
Widespread Adoption in China
Beyond Baidu, other major Chinese tech companies like Alibaba and Tencent are also making their AI models free and open-source. Alibaba Cloud, for example, announced plans to open-source its AI model for video creation, while Tencent recently released five new open-source models that convert text and images into 3D visuals.
Smaller players are also joining this trend. ManusAI, a Chinese AI company, announced its transition to open-source, aiming to give back to the AI community. Zhipu AI, one of China's leading AI startups, even declared that 2025 will be the "year of open-source."
According to Ray Wang, principal analyst at Constellation Research, DeepSeek’s emergence has made it impossible for competitors in China to sustain the same paid business model. Companies must adapt to open-source strategies to remain relevant.
Global Implications and Competition with the U.S.
DeepSeek’s success and China’s open-source AI movement have significant implications for global competition, particularly with U.S.-based AI companies like OpenAI and Anthropic. OpenAI, which sparked the AI revolution by launching ChatGPT in 2022, continues to operate under a proprietary model and has shown no signs of shifting to open-source.
This contrast in strategy questions about the effectiveness of massive investments in closed AI development. Microsoft, for instance, has invested $13 billion in OpenAI and is currently in discussions to raise an additional $40 billion in funding.
In contrast China’s open-source strategy has reduced AI development costs and expanded innovation opportunities. According to Tim Wang, managing partner at hedge fund Monolith Management, open-source AI models like DeepSeek’s have enabled breakthroughs with fewer resources.
The Future of Open-Source AI
While debates continue over whether China’s open-source approach is superior to the proprietary models in the U.S., many experts agree that this trend underscores the power of collaboration and openness in AI innovation. Alibaba Group Chairperson Joe Tsai emphasized that DeepSeek’s success is not just about competition between China and the U.S., but about how open-source AI can empower individuals and businesses to develop broader AI applications.
As more companies adopt open-source strategies, the future of AI appears increasingly inclusive and collaborative, fostering faster innovation worldwide.
Conclusion
China's adoption of open-source AI has reshaped the global AI industry. DeepSeek’s success has prompted major companies like Baidu, Alibaba, and Tencent to open their AI models, creating a broader and more inclusive innovation ecosystem.
This open-source approach not only benefits China’s AI industry but also lowers development costs and accelerates AI adoption across various sectors. While the U.S. maintains its proprietary AI business model, China is taking a different path, aiming to dominate the AI industry through openness and collaboration.
Ultimately, the long-term success of this strategy will be tested by evolving market dynamics and global regulations. However, one thing is clear—open-source AI has become a transformative force, redefining how AI technology is developed and adopted worldwide.
Read More : Manus: The Autonomous AI Revolution from China That Changes Everything